Early Intra-articular Corticosteroid Injection Improves Pain and Function in Adhesive Capsulitis of the Shoulder: 1-Year Retrospective Longitudinal Study

Early Intra-articular Corticosteroid Injection Improves Pain and Function in Adhesive Capsulitis of the Shoulder: 1-Year Retrospective Longitudinal Study

Accepted Manuscript Early intra-articular corticosteroid injection improves pain and function in adhesive capsulitis of shoulder: 1-year retrospective...

707KB Sizes 0 Downloads 29 Views

Accepted Manuscript Early intra-articular corticosteroid injection improves pain and function in adhesive capsulitis of shoulder: 1-year retrospective longitudinal study Jung Hwan Ahn, MD, Doo-Hyung Lee, MD, PhD, Hyuncheol Kang, PhD, Michael Y. Lee, MD, MHA, Dae Ryong Kang, PhD, Seung-Hyun Yoon, MD, PhD PII:

S1934-1482(17)30662-7

DOI:

10.1016/j.pmrj.2017.06.004

Reference:

PMRJ 1917

To appear in:

PM&R

Received Date: 27 June 2016 Revised Date:

25 May 2017

Accepted Date: 3 June 2017

Please cite this article as: Ahn JH, Lee D-H, Kang H, Lee MY, Kang DR, Yoon S-H, Early intraarticular corticosteroid injection improves pain and function in adhesive capsulitis of shoulder: 1-year retrospective longitudinal study, PM&R (2017), doi: 10.1016/j.pmrj.2017.06.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Title: Early intra-articular corticosteroid injection improves pain and function in adhesive

RI PT

capsulitis of shoulder: 1-year retrospective longitudinal study

Authors: Jung Hwan Ahn, MD, Doo-Hyung Lee, MD, PhD, Hyuncheol Kang, PhD, Michael Y. Lee, MD, MHA, Dae Ryong Kang, PhD, Seung-Hyun Yoon, MD, PhD

Affiliation: Department of Emergency Medicine (J.H. Ahn), Department of Orthopedic

SC

Surgery (D.-H. Lee), Medical Humanities and Social Medicine (D.R. Kang), and Physical Medicine and Rehabilitation (S.-H. Yoon), Ajou University School of Medicine, Suwon,

M AN U

Republic of Korea

Department of Applied Statistics (H. Kang), Hoseo University, Asan, Republic of Korea Department of Physical Medicine and Rehabilitation (M.Y. Lee), University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, NC

Correspondence: Seung-Hyun Yoon, MD, PhD, Department of Physical Medicine and

TE D

Rehabilitation, Ajou University School of Medicine, 164, World Cup-ro, Yeongtong-gu, Suwon 16499, Republic of Korea E-mail: [email protected]

Telephone: +82-31-219-5279

EP

Fax: +82-31-219-5209

AC C

Conflict of interest: none declared. This manuscript contains original material only that has not previously been published, and is not currently under consideration elsewhere, nor will be submitted elsewhere until a final decision has been made by the journal.

ACCEPTED MANUSCRIPT

Title: Early intra-articular corticosteroid injection improves pain and function in

2

adhesive capsulitis of shoulder: 1-year retrospective longitudinal study

AC C

EP

TE D

M AN U

SC

RI PT

1

1

ACCEPTED MANUSCRIPT

Background: Intra-articular corticosteroid injection is a commonly used therapy for

4

adhesive capsulitis, but not enough studies exist on the optimal timing of the injection.

5

Objective: To determine whether intra-articular corticosteroid injection has better

6

outcomes in patients with earlier stage than later stage of adhesive capsulitis.

7

Study design: Retrospective longitudinal study

8

Setting: University-affiliated tertiary care hospital.

9

Participants: Primary adhesive capsulitis patients (n=339) who were unresponsive to at

10

least 1 month of conservative treatment and who had ultrasound-guided corticosteroid

11

injection.

12

Interventions: Not applicable.

13

Main Outcome Measurements: Visual analog scale, Shoulder Pain and Disability Index,

14

and passive range of motion (flexion, abduction, external rotation, and internal rotation

15

and extension) were evaluated at pretreatment, month 1 and 12 after the first injection.

16

Results: The result of the multiple regressions, which considered the main and the

17

interaction effect of confounding variables, showed that the differences of all outcomes

18

in both short-term effect at month 1 and long-term effect at month 12 are greater when

19

the duration of pain prior to injection is shorter. Among the confounders, the injection

20

number in the difference of internal rotation and extension between month 0 and 12

21

(IRE ∆(0-12)) was statistically significant. IRE ∆(0-12) was also greater when the pain

22

duration was shorter, though the decrease in IRE ∆(0-12) differed depending on the

23

number of injections.

24

Conclusions: Early injection improves outcomes of adhesive capsulitis at both short-

25

and long-term follow-ups. If pain persists despite non-invasive and conservative

26

treatments, early injection may be considered to shorten its natural history.

27

Keywords: periarthritis; shoulder pain; treatment outcome; prognosis; intra-articular

28

injections; adhesive capsulitis; frozen shoulder

AC C

EP

TE D

M AN U

SC

RI PT

3

2

ACCEPTED MANUSCRIPT

Introduction

30

Adhesive capsulitis is a long-lasting, painful shoulder condition commonly known as

31

frozen shoulder or painful stiff shoulder. The condition results from progressive fibrosis

32

and eventual contracture of the capsule of the glenohumeral joint, which causes pain

33

and stiffness [1,2]. The treatment for adhesive capsulitis includes supervised exercise,

34

physical modalities, nonsteroidal anti-inflammatory drugs, oral corticosteroid, intra-

35

articular corticosteroid injection (IACI), suprascapular nerve block, hydraulic

36

arthrographic capsular distension, manipulation and arthroscopic capsular release [1-8].

37

IACI is one of the conservative treatments for adhesive capsulitis. It is effective in

38

rapidly reducing the pain[3], but there is a paucity of studies that define the optimal

39

timing of the injection. It is in the early stages of adhesive capsulitis that a

40

hypervascular synovial hyperplasia is present and this may eventually lead to the

41

fibrosis of the capsule. It has been contended that if given at an early (e.g. freezing)

42

stage, IACI may possibly minimize the morbidity of the disorder. Early treatment with

43

IACI may provide a chemical ablation of the synovitis which can help prevent the

44

development of the fibrosis and shorten the natural history of the disease [2,4,9,10].

45

However, not enough studies exist to prove the effectiveness of early injection. The aim

46

of this study is to evaluate whether early IACI is beneficial for pain and function.

47

TE D

M AN U

SC

RI PT

29

Methods

49

Patients

50

This is a 1-year retrospective longitudinal study based on the shoulder disease registry

51

data from the university-affiliated tertiary care hospital. Informed consent was waived

52

for this retrospective study as part of an exemption approved by the institutional review

53

board. 661 patients diagnosed with primary adhesive capsulitis were reviewed

54

retrospectively between January 1st, 2011 and December 31st, 2014. They were

55

outpatients at rehabilitation and orthopedic clinics. All the patients underwent a

56

standardized history-taking, physical examination and ultrasonographic evaluation by

57

the lead author; passive and active range of motion, painful arc and impingement test,

58

resisted test, strength of muscles in the affected shoulder at the initial workup.

59

Inclusion criteria were patients who had adhesive capsulitis with a normal radiograph

60

finding of the affected shoulder and restriction of passive motion of greater than 30˚ in 2

AC C

EP

48

3

ACCEPTED MANUSCRIPT

or more planes of movement by goniometric measurement with scapula rotation

62

restrained [11,12]; pain duration ranging from 3 to 12 months (freezing and frozen stage

63

of adhesive capsulitis); and average pain intensity during a week defined as a score of 6

64

points or more on a 10-cm visual analog scale (VAS) rated from “no pain” to “the worst

65

imaginable pain”. Abnormal limits of passive motion were defined as forward flexion

66

<150˚, extension <20˚, abduction <150˚, external rotation <60˚, and internal rotation

67

<60˚. External rotation and internal rotation were measured at shoulder 90˚ abduction. If

68

abduction measured was less than 90°, maximum possible abduction was achieved

69

before measuring the external rotation. To restrict the contribution of scapular rotation,

70

the patient was asked to relax, and the evaluator held the clavicle and scapular spine

71

with the hand. Passive elevation of the upper arm was stopped, and goniometric angles

72

were taken when it was no longer possible to prevent the scapula from rotating [13].

73

Patients were excluded if they had any of the following: secondary adhesive capsulitis

74

(secondary to other causes including inflammatory or infectious arthritis, stroke, tumor,

75

trauma, or fracture); history of diabetes and/or thyroid dysfunction; rotator cuff disease

76

on both physical and ultrasonographic examinations; medium to large sized (≥3cm)

77

partial- or full-thickness tear of the rotator cuff on ultrasonography [14,15]; calcific

78

tendinitis; primary osteoarthritis of the glenohumeral joint in a simple radiograph; and

79

previous corticosteroid injection on the affected shoulder. Magnetic resonance imaging

80

was not undertaken routinely, but those with either muscle weakness or a sign of

81

positive impingement were excluded on the basis of a high likeliness of rotator cuff tear.

82

A minimum of 1 month conservative treatment was provided without any IACI. The

83

conservative treatment included analgesics (nonsteroidal anti-inflammatory drugs,

84

acetaminophen/tramadol, opioids, tricyclic antidepressants, and fentanyl patch),

85

flexibility exercise, and physical modalities (heat and electrical therapies). Patients

86

received an institutional flexibility exercise of glenohumeral joint with physical

87

therapists once or two times a week for 2-4 weeks, and were educated to do the same

88

exercise at home. Patients who complained about persistent shoulder pain of VAS≥6

89

even after 1 month of conservative treatment were recommended to take the IACI.

AC C

EP

TE D

M AN U

SC

RI PT

61

90 91

Ultrasound-guided corticosteroid injection and exercise education

92

Patients sat in a seated position with their hands on their thighs. After aseptic 4

ACCEPTED MANUSCRIPT

preparation, the lead author inserted a 23-gauge, 6 cm long needle parallel to the

94

transducer from the posterior side of the shoulder. The needle was advanced under a

95

real-time ultrasound equipment (Logiq P6, GE Healthcare, Buckinghamshire, United

96

Kingdom) using 10-12MHz, linear array transducers until the needle tip entered the

97

glenohumeral joint. The lead author injected 2mL of 10mg/mL triamcinolone acetonide

98

and 8mL of 1% lidocaine. After injection, the patients were instructed about the home

99

exercise program for increasing the range of motion including stretching forward and

100

bending down to a desk, Codman exercise, wall-climbing exercise, external and internal

101

rotation with bar, and posterior shoulder stretch [16]. Exercises were to be performed 3

102

times a day lasting 10 minutes each round. No nonsteroidal anti-inflammatory drugs,

103

analgesics, or opioids were prescribed after injection. Patients were asked to visit the

104

hospital again at month 1, 2, and 3 after the injection and recommended to have a

105

second or third injection if their pain persisted over 50%.

106

M AN U

SC

RI PT

93

Outcome measurements

108

VAS, Shoulder Pain and Disability Index score (SPADI), passive range of motion were

109

compared for pre-, 1 month (short-term effect of IACI), and 12 month (long-term effect)

110

after the first injection. All outcome measurements were evaluated by the lead author. A

111

VAS as primary outcome measurement was used to assess shoulder pain. SPADI is a

112

self-reporting questionnaire for patients with shoulder pain, which consists of 13

113

questions that are divided into 2 domains: pain (5 items) and disability (8 items) [17].

114

Each domain score is equally weighted and added to obtain a total percentage score

115

between the ranges of 0 (best) to 100 (worst). Passive range of motion of forward

116

flexion, abduction and external rotation were measured by goniometer in supine

117

position. External rotation was measured in 0° abduction of shoulder and 90° flexion of

118

elbow position. Internal rotation and extension was measured as the distance (cm) from

119

the spinous process of 7th cervical vertebra to the tip of the fully extended thumb in

120

standing position [18]. This methodology shows high intrarater reliability [19] and is a

121

useful method that is also relevant with the activities of daily living, such as dressing,

122

bathing, and toileting [20].

AC C

EP

TE D

107

123 124

Statistical analysis 5

ACCEPTED MANUSCRIPT

All the data were analyzed by using the SPSS version 23 statistics software (IBM,

126

Armonk, NY). Continuous data are expressed as means ± standard deviation.

127

Categorical data are shown as absolute values, together with the frequency distribution.

128

The number of injections were reclassified into two groups, single-injection (injection

129

number 1) group and double- or triple-injection (injection number 2 or 3) group; the two

130

latter were combined since only 7 (2.1%) out of 339 patients opted for triple-injection.

131

The multiple regression analysis was conducted to determine the relationship between

132

the differences of outcomes and pain duration while controlling the confounding

133

variables (gender, age, injection number, or arm dominance). Simple regression was

134

also conducted between the differences of outcomes and other variables (age, gender,

135

the group according to injection number, arm dominance or pain duration) in order to

136

select the variables which were put into multiple regression. The variables with P ≤ .05

137

in the simple regression were entered into the multiple regression with main and

138

interaction effects. P <.05 was considered significant in multiple regression analysis. In

139

the multiple regression analysis, it was considered significant if P <.05.

M AN U

SC

RI PT

125

140

Results

142

General characteristics

143

Medical records of a total of 661 patients were reviewed. Of them, 246 had pain that

144

improved after conservative treatment (figure 1). IACI was recommended to the rest of

145

the patients (415) with pain of VAS ≥6; 40 refused and 375 accepted the

146

recommendation. Of the 375 who accepted, 36 were not included in the 12-month

147

follow-up; 17 did not attend the follow-up, 8 were treated at other clinics, and 11 could

148

not be reached. Finally, 339 patients completed the follow-up at month 1 (short-term

149

effects) and 12 (long-term) after IACI. The mean (range) age was 54.5±6.0 (39 to 69)

150

years old; ratio of male to female 111 (32.7%): 228 (67.3%); ratio of dominant to non-

151

dominant arm 129 (38.1%): 210 (61.9%); pain duration 8.2±3.0 (3 to 12) months; and

152

ratio of injection (single- versus double- and triple-injection) 146 (43.1%): 193 (56.9%).

153

Outcome measurements and complications

154

Table 1 shows the value and changes of outcome measurements after IACI. Apart from

155

facial flushing on days 2 to 6 after injection (12 patients), there were no serious

156

complications such as infection.

AC C

EP

TE D

141

6

ACCEPTED MANUSCRIPT

Selection of confounders

158

Table 2 shows the results of simple regression analysis. In case of short-term effects, or

159

the difference in month 0 to 1 (∆(0-1) or ∆(1-0)), arm dominance was the confounder in

160

only VAS ∆(0-1), while no confounders were found in other outcomes (SPADI ∆(0-1),

161

flexion ∆(1-0), abduction ∆(1-0), external rotation ∆(1-0), internal rotation and

162

extension ∆(0-1)). In long-term effects, or the difference in month 0 to 12 (∆(0-12) or

163

∆(12-0)), the confounders were injection number and arm dominance in VAS ∆(0-12),

164

injection number and gender in SPADI ∆(0-12), while no confounder was found in

165

flexion ∆(12-0), injection number and gender in abduction ∆(12-0), injection number in

166

external rotation ∆(12-0), and injection number and gender in internal rotation and

167

extension ∆(0-12) (P <.05).

168

Relation between the differences of outcomes and pain duration controlling confounders

169

Table 3 shows the results of multiple regression for short-term effects. Pain duration

170

was statistically significant in all the differences of outcomes (flexion, P =.016; others,

171

P <.001). Beta (regression coefficient) of pain duration in all the differences of

172

outcomes was negative. Arm dominance as confounder in VAS ∆(0-1) was not

173

statistically significant considering the main and the interaction effect with pain

174

duration (P >.05). Table 4 shows the results of multiple regression for long-term effects.

175

Pain duration was statistically significant in all the differences of outcomes (flexion, P

176

=.001; others, P <.001). Beta of pain duration in all the differences of outcomes was

177

negative. Among the confounders in the differences of outcomes, the main and the

178

interaction effect of injection number in IRE ∆(0-12) was statistically significant

179

(injection number, P =.002; interaction between pain duration and injection number, P

180

=.001), while the other confounders were not statistically significant. The slopes which

181

represent the different injection number groups were negative, x axis being pain

182

duration and y axis IRE ∆(0-12) (figure 2).

SC

M AN U

TE D

EP

AC C

183

RI PT

157

184

Discussion

185

This 1-year retrospective longitudinal study reveals that the improvement in outcomes,

186

which considered the main and the interaction effect of the confounding variables, are

187

greater when the pain duration prior to IACI is shorter in both short-term effect at

188

month 1 and long-term effect at month 12. Among the confounders in the improvement 7

ACCEPTED MANUSCRIPT

in outcomes, only the injection number in IRE ∆(0-12) resulted in a statistically

190

significant confounder resulting in better outcomes with more injections. Though the

191

decrease in IRE ∆(0-12) differed depending on the number of injections, IRE ∆(0-12)

192

was also greater when the pain duration was shorter. These results suggest that early

193

IACI will have a greater improvement in pain and function than late IACI.

194

Many researchers have dismissed adhesive capsulitis as a benign condition, with most

195

being self-resolving [1,21]. However, it is now widely accepted that up to 50% of the

196

patients may not fully recover from the symptoms and disability, even with long-term

197

follow-ups [22-24]. Some investigators have advocated the use of IACI in the treatment

198

of early adhesive capsulitis and suggested that patients treated with injection at an early

199

stage of the disorder respond better than later stage. Their rationale for early IACI is that

200

it would reduce inflammation of synovium, decreasing capsular fibrosis and shortening

201

the natural history of the disease [2,4,9,10,25,26]. However, not enough studies exist to

202

prove the effects of early injection. In a previous retrospective case series, the authors

203

showed that IACI may be more efficacious in the freezing stage (stage 1 and 2 based on

204

the classification by Neviaser and Neviaser’s four staging system [27]) before the

205

development of a significant capsular contracture [25]. In the same study, the result

206

showed that an early IACI for adhesive capsulitis may have allowed patients to recover

207

motion at a median time of 3 months. But the study had several limitations, including

208

patients in stage 1 (n=5) who would be difficult to diagnose if not for the support of an

209

arthroscopic or histologic evaluation; no control group of frozen stage (stage 3 and 4)

210

patients to compare the effectiveness of IACI; relatively few samples (n=44 in total);

211

and a short mean follow-up period of 6.7 months. Our study overcame previous

212

limitations by including frozen stage patients, a relatively large sample (n=339), and a

213

long enough follow-up (1-yr) to observe the effect of IACI.

214

The natural history of adhesive capsulitis is commonly described in three stages [4,5,21];

215

freezing, frozen and thawing stages. The first or the freezing stage is marked by

216

progressive pain and loss of motion lasting up to 9 months. The second or the frozen

217

stage lasts from 4 to 12 months. In this stage, pain improves but stiffness persists and

218

continues to impact patients’ ability to perform activities of daily living. The third and

219

final thawing stage is a period of recovery marked by gradual improvement in the range

220

of motion that can take up to 12 to 42 months. Since the purpose of IACI is to reduce

AC C

EP

TE D

M AN U

SC

RI PT

189

8

ACCEPTED MANUSCRIPT

pain caused by capsular inflammation, many authors claim that IACI is effective in

222

freezing or early frozen stages with moderate to severe pain [2,28]. We included patients

223

with pain duration of up to 12 months, which is considered to be long enough to include

224

patients with freezing to frozen stages.

225

IACI is considered to be a well-known and accepted treatment for adhesive capsulitis.

226

Based on the concept of synovial inflammation for the pathology of adhesive capsulitis,

227

the joint capsule can be the main pathology, and IACI can be justified for its treatment

228

[29]. However, not enough evidences exist to conclude which injection method (intra-

229

articular versus subacromial injection) is superior for the treatment of adhesive

230

capsulitis. There are three studies that compared the location of injection. Oh et al [29]

231

found that an intra-articular injection did not have a significantly greater decrease in the

232

VAS score compared with a subacromial injection. Another study reported similar

233

improvements in range of motion between intra-articular and subacromial injections

234

[30]. A more recent study compared the efficacy of 3 injection methods, subacromial

235

injection, intra-articular injection, and hydrodilatation in the treatment of primary

236

adhesive capsulitis. Although hydrodilatation yielded more rapid improvement, the 3

237

injection methods resulted in similar improvement in the final follow-up at 6 months

238

[31].

239

Ultrasound was used as guidance for IACI in this study. Traditionally, IACI for the

240

treatment of adhesive capsulitis has been performed guided by anatomical landmarks

241

alone. With the advent of readily available imaging tools such as ultrasound, image-

242

guided injections have increasingly become accepted into routine care. While there is

243

some evidence that the use of imaging improves accuracy, it is unclear from current

244

evidence whether or not it improves patient-relevant outcomes [32]. Lee et al [33]

245

compared ultrasound-guided injections of corticosteroid with blind injections for

246

adhesive capsulitis during 6 weeks. The improvement in pain intensity, range of motion,

247

and shoulder function score was significantly greater in the ultrasound-guided injection

248

group than in the blind injection group by the second week post-injection. However,

249

there were no further significant differences in the improvement between the 2 groups

250

beyond the third week. Raeissadat et al [34] also compared ultrasound-guided injections

251

with blind injections for a 4-week period. Improvements in pain intensity, range of

252

motion, and functional score after 1 and 4 weeks were more prominent in the

AC C

EP

TE D

M AN U

SC

RI PT

221

9

ACCEPTED MANUSCRIPT

ultrasound-guided group, but the differences were not statistically significant, except for

254

the changes in extension where the improvements were significantly higher in the

255

ultrasound-guided group. The accuracy of injections was also higher in the ultrasound-

256

guided group (90 % vs. 76 %), but the differences were not found to be significant.

257

Overall, ultrasound-guided injections in adhesive capsulitis are more accurate. However,

258

there were either no statistically significant correlations between the accuracy of

259

injection and improvement of patients’ symptoms [34], or, even if found, the differences

260

were insignificant after the third week [33].

261

In this study, we used a sufficient amount of local anesthetic (8 mL of 1% lidocaine) to

262

reduce pain during and after the injection. Several in vitro studies reported the

263

chondrotoxic effect of local anesthetics [35,36]. And another study reported that the

264

osteoarthritic articular cartilage is more vulnerable to local anesthetics than normal

265

articular cartilage [37]. However, a more recent study investigated the in vivo effects of

266

local anesthetics on the articular cartilage of the knees in rats. The experiment was first

267

performed on normal knees, and then repeated on joints affected by experimentally

268

induced osteoarthritis by transection of the anterior cruciate ligament. Results showed

269

that intra-articular injections with 10 mL of 0.5% bupivacaine repeated once a week for

270

5 consecutive weeks did not have a significant effect on the viability and density of

271

chondrocytes, nor on the histological features of articular cartilage when compared with

272

saline solution injections. The effects of local anesthetic were nonsignificant at different

273

time points, regardless of whether osteoarthritic changes were induced [38]. However,

274

these effects are limited to 0.5% bupivacaine, and further studies are needed for other

275

types of local anesthetics. Since the toxic effect to chondrocyte of local anesthetics is

276

still under investigation, it would be appropriate to use as little anesthetic volume as

277

possible.

278

We adopted widely-applied diagnostic criteria of adhesive capsulitis, commonly used

279

also in previous studies [11,12,39,40]. According to this criteria, a patient will be

280

diagnosed as adhesive capsulitis when his/her range of motion is decreased more than

281

30˚ on 2 planes based on normal range of motion (forward flexion 180˚, extension 50˚,

282

abduction 180˚, external rotation 90˚, internal rotation 90˚). Primary adhesive capsulitis

283

is caused by global capsular inflammation and fibrosis, which results in a global

284

limitation of range of motion. Recent studies have, however, argued that an

AC C

EP

TE D

M AN U

SC

RI PT

253

10

ACCEPTED MANUSCRIPT

anterosuperior or anteroinferior capsular tightening is another characteristic [6] and

286

included a limitation of external rotation in the diagnostic criteria of adhesive capsulitis;

287

external rotation to less than one half of normal [28] or less than 20° of external rotation

288

[13,41]. Considering these changes, it would be more ideal to use a diagnostic criteria

289

that would include the limitation of external rotation in the future studies on adhesive

290

capsulitis.

291

Longer pain duration will most probably push the number of injections up, which can be

292

another reason for advocating early injection. We conducted an additional statistical

293

analysis to explore the relationship between pain duration and the number of injections.

294

The result showed a trend of longer pain leading to increased number of injections.

295

However, factors (dominant arm, gender, and VAS 1) other than pain duration should be

296

considered before drawing any conclusions based merely on injection counts. Further

297

study that considers other factors (e.g. socioeconomic status or personal preference for

298

injection et al) not included in this study will be needed to clarify the correlation

299

between pain duration and number of injections.

300

The greatest limitation of this study lies in the fact that it is a retrospective study.

301

Nonetheless, it has a low drop-out rate (9.6% in total) and includes at least 4 rounds of

302

follow-ups (at month 1, 2, 3, and 12) after IACI to monitor any patients that have

303

received other treatments in a great effort to reduce bias. Second, we excluded patients

304

with diabetes or thyroid dysfunction. However, adhesive capsulitis is often associated

305

with diabetes or thyroid dysfunction [42,43]; the presence of these diseases may be an

306

important prognostic factor for adhesive capsulitis. Therefore, it would be necessary to

307

have further research that would include these patients. Third, this study included

308

patients with pain duration ranging from 3 to 12 months. However, the duration of pain

309

was determined entirely based on the patient’s memory, possibly leading to a recall bias.

SC

M AN U

TE D

EP

AC C

310

RI PT

285

311

Conclusion

312

This 1-year follow-up retrospective study reveals that an early IACI shows greater

313

improvement in pain and function than late IACI. If pain persists despite non-invasive

314

and conservative treatments, an early injection may be considered to shorten its natural

315

history.

11

ACCEPTED MANUSCRIPT

Acknowledgment

317

The authors thank Aeree Park, MA, for the English translation of the Korean manuscript.

AC C

EP

TE D

M AN U

SC

RI PT

316

12

ACCEPTED MANUSCRIPT

318

References

319

1.

320

Orthop Surg 2011;19:536-542. 2.

322 323

Hannafin JA, Chiaia TA. Adhesive capsulitis. A treatment approach. Clin Orthop Relat Res 2000;:95-109.

3.

324

RI PT

321

Neviaser AS, Neviaser RJ. Adhesive capsulitis of the shoulder. J Am Acad

Buchbinder R, Green S, Youd JM. Corticosteroid injections for shoulder pain. Cochrane Database Syst Rev 2003;:CD004016.

4.

Dias R, Cutts S, Massoud S. Frozen shoulder. BMJ 2005;331:1453-1456.

326

5.

Chambler AF, Carr AJ. The role of surgery in frozen shoulder. J Bone Joint Surg

327

Br 2003;85:789-795. 6.

329 330

Robinson CM, Seah KT, Chee YH, Hindle P, Murray IR. Frozen shoulder. J

M AN U

328

SC

325

Bone Joint Surg Br 2012;94:1-9. 7.

Karatas GK, Meray J. Suprascapular nerve block for pain relief in adhesive

331

capsulitis: comparison of 2 different techniques. Arch Phys Med Rehabil

332

2002;83:593-597.

333

8.

Lee DH, Yoon SH, Lee MY, Kwack KS, Rah UW. Capsule-preserving hydrodilatation with corticosteroid versus corticosteroid injection alone in

335

refractory adhesive capsulitis of shoulder: a randomized controlled trial. Arch

336

Phys Med Rehabil 2016.

337

9.

TE D

334

Walmsley S, Osmotherly PG, Walker CJ, Rivett DA. Power Doppler ultrasonography in the early diagnosis of primary/idiopathic adhesive capsulitis:

339

an exploratory study. J Manipulative Physiol Ther 2013;36:428-435. 10.

341 342

shoulder. J Bone Joint Surg Br 2007;89:928-932.

11.

343

with painful stiff shoulder: results of a randomised, double blind, placebo

345

controlled trial. Ann Rheum Dis 2004;63:302-309. 12.

347 348 349

Buchbinder R, Green S, Forbes A, Hall S, Lawler G. Arthrographic joint

distension with saline and steroid improves function and reduces pain in patients

344

346

Hand GC, Athanasou NA, Matthews T, Carr AJ. The pathology of frozen

AC C

340

EP

338

Warner JJ. Frozen Shoulder: Diagnosis and Management. J Am Acad Orthop Surg 1997;5:130-140.

13.

Koh ES, Chung SG, Kim TU, Kim HC. Changes in biomechanical properties of glenohumeral joint capsules with adhesive capsulitis by repeated capsule13

ACCEPTED MANUSCRIPT

350

preserving hydraulic distensions with saline solution and corticosteroid. PM R

351

2012;4:976-984. 14.

353

Relat Res 1990:64-74.

355 356

initial rotator-cuff repair. J Bone Joint Surg Am 1984;66:563-567. 16.

357 358

Matsen FA, Lippitt S, Sidles J, Harryman DT. Practical evaluation and management of the shoulder: Saunders Philadelphia:; 1994, 45-49.

17.

359 360

DeOrio JK, Cofield RH. Results of a second attempt at surgical repair of a failed

RI PT

15.

Roach KE, Budiman-Mak E, Songsiridej N, Lertratanakul Y. Development of a

SC

354

Ellman H. Diagnosis and treatment of incomplete rotator cuff tears. Clin Orthop

shoulder pain and disability index. Arthritis Care Res 1991;4:143-149. 18.

Edwards TB, Bostick RD, Greene CC, Baratta RV, Drez D. Interobserver and

M AN U

352

361

intraobserver reliability of the measurement of shoulder internal rotation by

362

vertebral level. J Shoulder Elbow Surg 2002;11:40-42.

363

19.

364 365

Hoving JL, Buchbinder R, Green S, et al. How reliably do rheumatologists measure shoulder movement? Ann Rheum Dis 2002;61:612-616.

20.

Rah UW, Yoon SH, Moon DJ, et al. Subacromial corticosteroid injection on poststroke hemiplegic shoulder pain: a randomized, triple-blind, placebo-

367

controlled trial. Arch Phys Med Rehabil 2012;93:949-956. 21.

369

Rheumatol 1975;4:193-196. 22.

371 372

421. 23.

373 374

Shaffer B, Tibone JE, Kerlan RK. Frozen shoulder. A long-term follow-up. J Bone Joint Surg Am 1992;74:738-746.

24.

375 376

Hazleman BL. The painful stiff shoulder. Rheumatol Phys Med 1972;11:413-

EP

370

Reeves B. The natural history of the frozen shoulder syndrome. Scand J

AC C

368

TE D

366

Binder AI, Bulgen DY, Hazleman BL, Roberts S. Frozen shoulder: a long-term prospective study. Ann Rheum Dis 1984;43:361-364.

25.

Marx RG, Malizia RW, Kenter K, Wickiewicz TL, Hannafin JA. Intra-articular

377

corticosteroid injection for the treatment of idiopathic adhesive capsulitis of the

378

shoulder. HSS J 2007;3:202-207.

379

26.

380 381

Bunker TD, Anthony PP. The pathology of frozen shoulder. A Dupuytren-like disease. J Bone Joint Surg Br 1995;77:677-683.

27.

Neviaser RJ, Neviaser TJ. The frozen shoulder. Diagnosis and management. Clin 14

ACCEPTED MANUSCRIPT

382 383

Orthop Relat Res 1987;:59-64. 28.

384 385

Hsu JE, Anakwenze OA, Warrender WJ, Abboud JA. Current review of adhesive capsulitis. J Shoulder Elbow Surg 2011;20:502-514.

29.

Oh JH, Oh CH, Choi JA, Kim SH, Kim JH, Yoon JP. Comparison of glenohumeral and subacromial steroid injection in primary frozen shoulder: a

387

prospective, randomized short-term comparison study. J Shoulder Elbow Surg

388

2011;20:1034-1040. 30.

390 391

Rizk TE, Pinals RS, Talaiver AS. Corticosteroid injections in adhesive capsulitis: investigation of their value and site. Arch Phys Med Rehabil 1991;72:20-22.

31.

SC

389

RI PT

386

Yoon JP, Chung SW, Kim JE, et al. Intra-articular injection, subacromial injection, and hydrodilatation for primary frozen shoulder: a randomized clinical

393

trial. J Shoulder Elbow Surg 2016;25:376-383.

394

32.

M AN U

392

Bloom JE, Rischin A, Johnston RV, Buchbinder R. Image-guided versus blind

395

glucocorticoid injection for shoulder pain. Cochrane Database Syst Rev

396

2012:CD009147.

397

33.

Lee HJ, Lim KB, Kim DY, Lee KT. Randomized controlled trial for efficacy of intra-articular injection for adhesive capsulitis: ultrasonography-guided versus

399

blind technique. Arch Phys Med Rehabil 2009;90:1997-2002.

400

34.

TE D

398

Raeissadat SA, Rayegani SM, Langroudi TF, Khoiniha M. Comparing the accuracy and efficacy of ultrasound-guided versus blind injections of steroid in

402

the glenohumeral joint in patients with shoulder adhesive capsulitis. Clin

403

Rheumatol 2016. 35.

405 406

bupivacaine on articular chondrocytes. J Bone Joint Surg Br 2008;90:814-820.

36.

407

Arthroscopy 2009;25:707-715.

37.

410 411

Lo IK, Sciore P, Chung M, et al. Local anesthetics induce chondrocyte death in bovine articular cartilage disks in a dose- and duration-dependent manner.

408 409

Chu CR, Izzo NJ, Coyle CH, Papas NE, Logar A. The in vitro effects of

AC C

404

EP

401

Chu CR, Izzo NJ, Papas NE, Fu FH. In vitro exposure to 0.5% bupivacaine is cytotoxic to bovine articular chondrocytes. Arthroscopy 2006;22:693-699.

38.

Iwasaki K, Sudo H, Kasahara Y, et al. Effects of Multiple Intra-articular

412

Injections of 0.5% Bupivacaine on Normal and Osteoarthritic Joints in Rats.

413

Arthroscopy 2016;32:2026-2036. 15

ACCEPTED MANUSCRIPT

414

39.

Yoon SH, Lee HY, Lee HJ, Kwack KS. Optimal dose of intra-articular

415

corticosteroids for adhesive capsulitis: a randomized, triple-blind, placebo-

416

controlled trial. Am J Sports Med 2013;41:1133-1139. 40.

418 419

ed: Chicago, IL: AMA; 2012, 461-476. 41.

420 421

Rondinelli RD. AMA Guides to the Evaluation of Permanent Impairment. Sixth

RI PT

417

Ando A, Sugaya H, Hagiwara Y, et al. Identification of prognostic factors for the nonoperative treatment of stiff shoulder. Int Orthop 2013;37:859-864.

42.

Balci N, Balci MK, Tuzuner S. Shoulder adhesive capsulitis and shoulder range of motion in type II diabetes mellitus: association with diabetic complications. J

423

Diabetes Complications 1999;13:135-140.

425

43.

Wohlgethan JR. Frozen shoulder in hyperthyroidism. Arthritis Rheum 1987;30:936-939.

AC C

EP

TE D

426

M AN U

424

SC

422

16

ACCEPTED MANUSCRIPT

Figure 1. Flow diagram indicating progress of patients through the study.

428

IACI, intra-articular corticosteroid injection

AC C

EP

TE D

M AN U

SC

RI PT

427

17

ACCEPTED MANUSCRIPT

Figure 2. The multiple linear regression graph showing the difference of internal

430

rotation and extension between month 0 and 12 (IRE ∆(0-12)), pain duration, and the

431

groups according to the injection number. Dotted line represents the single-injection

432

group (injection number 1, IRE ∆(0-12) = 31.36 + (-1.40) x pain duration). Solid line

433

represents the double- or triple-injection group (injection number 2 or 3, IRE ∆(0-12) =

434

41.98 + (-2.67) x pain duration). IRE ∆(0-12) is greater when the pain duration is

435

shorter, though the decrease in the IRE ∆(0-12) differs depending on the number of the

436

injection.

AC C

EP

TE D

M AN U

SC

RI PT

429

18

ACCEPTED MANUSCRIPT Table 1. Changes of outcome measurements after intra-articular corticosteroid injection Month 1

Month 12

∆(0-1)

∆(0-12)

VAS score

7.6±1.0

3.2±1.4

1.9±1.3

4.4±1.8

5.7±1.7

SPADI score

59.5±5.6

30.6±11.0

24.1±9.9

28.9±12.5

35.5±11.4

Flexion, degree

144.9±18.7

153.4±18.2

166.4±13.5

Abduction, degree

72.4±14.7

115.7±33.8

148.9±33.8

ER, degree

32.4±10.8

40.5±8.0

73.5±15.8

IRE, cm

51.7±7.4

42.2±7.7

31.6±8.3

M AN U

∆(a-b) means difference of month a to b.

-8.5±14.3

-21.5±21.7

-43.3±34.0

-76.5±32.7

-8.1±11.1

-41.1±21.4

9.5±8.2

20.0±11.0

SC

Values are expressed as mean ± standard deviation.

RI PT

Month 0

VAS=visual analogue scale; SPADI=Shoulder Pain and Disability Index; ER=external

AC C

EP

TE D

rotation; IRE= internal rotation and extension.

ACCEPTED MANUSCRIPT

Table 2. Results of simple linear regressions. Gender

Beta

SE

P

Beta

SE

P

Beta

SE

∆(0-1)

-.33

.03

<.001

.01

.02

.759

-.35

.21

∆(0-12)

-.24

.03

<.001

-.01

.02

.681

.36

∆(0-1)

-1.84

.21

<.001

.01

.11

.963

-2.64

∆(0-12)

-1.68

.19

<.001

.05

.10

.632

∆(1-0)

-.63

.26

.016

.09

.13

.490

∆(12-0)

-1.35

.39

.001

.29

∆(1-0)

-5.16

.56

<.001

.11

∆(12-0)

-4.74

.54

<.001

∆(1-0)

-1.67

.18

∆(12-0)

-4.82

.29

∆(0-1)

-1.26

∆(0-12)

-2.06

Abduction

.24

AC C

ER

Beta

SE

P

Beta

SE

P

.089

-

-

-

.72

.20

<.001

.067

-1.21

.17

<.001

.37

.19

.047

1.45

.069

-

-

-

1.27

1.40

.368

-3.40

1.31

.010

-7.10

1.19

<.001

-.16

1.27

.902

-.43

1.67

.797

-

-

-

2.47

1.60

.124

SC .19

M AN U

TE D

Flexion

Arm dominance

P

VAS

SPADI

Injection number

RI PT

Age

EP

Pain duration

.20

.147

-.91

2.51

.718

-3.66

2.37

.124

3.97

2.42

.102

.31

.732

-5.50

3.93

.162

-

-

-

-1.24

3.81

.745

.30

.416

-7.46

3.77

.049

-20.91

3.41

<.001

4.87

3.66

.184

<.001

-.12

.10

.256

-2.24

1.28

.081

-

-

-

.99

1.24

.425

<.001

-.08

.20

.699

-1.77

2.48

.477

-16.69

2.17

<.001

2.87

2.39

.231

.13

<.001

-.10

.08

.173

-.86

.95

.369

-

-

-

-1.10

.92

.232

.17

<.001

-.06

.10

.570

-3.07

1.27

.016

-6.57

1.16

<.001

-.54

1.24

.664

IRE

ACCEPTED MANUSCRIPT

∆(a-b) means difference of month a to b.

AC C

EP

TE D

M AN U

SC

RI PT

VAS=visual analogue scale; SPADI=Shoulder Pain and Disability Index; ER= external rotation; IRE= internal rotation and extension.

ACCEPTED MANUSCRIPT Table 3. Short-term effects by multiple linear regressions. Response variables VAS ∆(0-1)

Effects Intercept *

Pain duration

B

SE

t

P

7.22

.29

24.56

<.001

-.30

.03

-90.3

<.001

-.15

.48

-.30

.763

Dominant Non-dominant

0

Pain duration x arm dominance Dominant

-.07

Non-dominant SPADI ∆(0-1)

Intercept

44.10

Pain duration Intercept

*

Intercept *

Pain duration ER ∆(1-0)

Intercept Pain duration*

IRE ∆(0-1)

Intercept

.

.06

-1.30

.196

.

.

.

1.81

24.41

<.001

.21

-8.91

<.001

13.64

2.28

5.99

<.001

-.63

.26

-2.42

.016

85.78

4.86

17.66

<.001

-5.16

.56

-9.30

<.001

21.86

1.59

13.72

<.001

-1.67

.18

-9.18

<.001

19.84

1.18

16.87

<.001

-1.26

.13

-9.35

<.001

TE D

Pain duration*

.

-1.84

M AN U

Pain duration Abduction ∆(1-0)

.

SC

*

Flexion ∆(1-0)

0

RI PT

Arm dominance

* The effects were statistically significant (P <.05).

EP

∆(a-b) means difference of month a to b.

VAS=visual analogue scale; SPADI=Shoulder Pain and Disability Index; ER=external

AC C

rotation; IRE=internal rotation and extension.

ACCEPTED MANUSCRIPT Table 4. Long-term effects by multiple linear regressions. Response variables VAS ∆(0-12)

Effects Intercept

Pain duration

*

B

SE

t

P

7.18

.50

14.34

<.001

-.19

.05

-3.77

<.001

.23

.55

.41

.681

1 2 or 3

0

.

.

.

.50

.17

.864

.06

-.80

.426

.04

.07

.68

.496

0

.

.

.

51.26

2.99

17.15

<.001

-2.09

.31

-6.75

<.001

-4.16

3.72

-1.12

.265

0

.

.

.

-2.31

3.49

-.66

.508

.40

1.68

.093

.78

.44

1.78

.076

0

.

.

.

Intercept

32.60

3.41

9.57

<.001

Pain duration*

-1.35

.39

-3.47

.001

120.71

8.67

13.92

<.001

-5.81

.90

-6.48

<.001

-13.04

10.80

-1.21

.228

0

.

.

.

-5.80

10.11

-.57

.567

Arm dominance Dominant

.07

Non-dominant

0

Dominant

-.05

Non-dominant

0

2 or 3 Intercept

Pain duration

*

Injection number 1

Gender

TE D

2 or 3

M AN U

Pain duration x injection number 1

SC

Pain duration x arm dominance

SPADI ∆(0-12)

RI PT

Injection number

Male

Female

0

Pain duration x gender

EP

Male

Female

.67 0

AC C

Pain duration x injection number 1

2 or 3

Flexion ∆(12-0)

Abduction ∆(12-0)

Intercept

Pain duration

*

Injection number 1 2 or 3 Gender Male

ACCEPTED MANUSCRIPT Female

0

Pain duration x gender Male

1.51

Female

1.15

1.31

.190

2.56

1.27

2.02

.045

0

.

.

.

4.63

17.63

<.001

.47

-10.64

<.001

5.79

-.63

.528

.

.

.

.73

.68

1.07

.284

0

.

.

.

41.98

2.68

15.69

<.001

-2.67

.28

-9.65

<.001

-10.62

3.33

-3.19

.002

0

.

.

.

4.22

3.12

1.35

.177

.35

-.31

.760

1.27

.39

3.25

.001

0

.

.

.

0

1 2 or 3 ER ∆(12-0)

Intercept

81.57

Pain duration*

-4.97

Injection number 1

-3.66 0

Pain duration x injection number 2 or 3 IRE ∆(0-12)

Intercept

Pain duration

*

Injection number 1* 2 or 3 Gender

M AN U

1

SC

2 or 3

RI PT

Pain duration x injection number

TE D

Male

Female

0

Pain duration x gender Male

Female

-.11 0

EP

Pain duration x injection number 1*

AC C

2 or 3

* The effects were statistically significant (P <.05). ∆(a-b) means difference of month a to b. VAS=visual analogue scale; SPADI=Shoulder Pain and Disability Index; ER=external rotation; IRE=internal rotation and extension.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT